Why stock market crashes happen in the autumn
Why do financial markets so often get conniptions in September? Why do so many market collapses happen in October?
Lily Fang, a professor at the Massachusetts Institute of Technology, has turned her attention to this nagging subject.
She blames summer holidays for the way stock market bubbles burst so frequently as autumn sweeps into the northern hemisphere.
In an amateur way, I have been saying something similar for years.
My explanation of why autumn is such a hazardous time has to do with the fact that investors are often buoyed up with hope when they go off on their summer holidays.
When they come back to their dealing desks in September, they began to look towards the end of the year… and realise that the small economic clouds they had been able to shrug off in midsummer are still there, and getting more stormy looking by the day.
That's when panic begins to set in.
Over and over again, in years such as 1929, 1987 and 2008, markets crash in the autumn.
Prof Fang has refined my gut feeling. What she brings to the subject is an up-to-date view of how the financial markets operate.
September is traditionally a bad time for stock and shares. It's the only month that shows a decline, on average, over the past 100 years, and over shorter spans of time.
It has been called the "September effect", but there has until recently been no real explanation for it.
Professor Fang and her colleagues think they know what causes it: the summer holidays mean that professional investors can't be as focussed on financial market news as they are for the rest of the year.
Instead they delay their reaction until they go back to work in September. And often this is bad news which takes longer to digest than good news, says the research.
Prof Fang and her colleagues have tested their thesis by looking at differences in school holiday dates in different places. and finding that the September effect varies depending on which month most people go away.
In other words, financial markets - so praised for their efficiency - get less efficient in the summer because people are not paying sufficient attention to what is going on.
Not rocket science this, but interesting.
There are other curious things about financial markets. One may or may not still be happening, but certainly was when I looked at it 30 years ago.
Long term studies of interest rates then detected a measurable tendency for them to rise in the northern hemisphere in the spring and fall in the late summer.
Why? Well researchers came to the conclusion that that is what happened to interest rates in an era when agriculture was the primary economic activity.
Farmers borrowed money from banks to plant crops in the spring and repaid the loans when they harvested the results of their planting in August.
In the 18th and 19th centuries, money was tight in the spring, and flowed back to the lenders later in the year.
The weird thing was that this was still a detectable trend in the 1980s - nearly 100 years after agriculture had lost its dominant grip on Western economies.
Financial markets certainly have "ghost memories". For decades in the hustling Chicago Board of Trade, much trading used to be done in the pits.
Professional traders spent their lives crowded into these small spaces, shouting prices at each other across the pit.
This open outcry trading was a very physical business. On busy days traders jostled for a space.
It is said that one trader who had a heart attack and died was kept upright by the crush of other market participants (who hadn't noticed), until the session came to an end.
When market prices were moving fast, some traders in agricultural commodities such as corn and wheat could find themselves physically unable to get deep enough into the pit to trade.
So they turned to the neighbouring less excitable pit where silver was traded.
The result was that when the price of farm commodities was rising fast in Chicago, silver would tend to go up at in a similar way, through sheer association.
In the 1980s the Board of Trade changed the layout of its trading pits so that silver was no longer adjacent to corn.
But for a long time afterwards, busy farm trading days tended to have a knock on effect on the silver market.
No longer because they were next to each other, but because that's the way it had happened for decades. There are limits to rationality.